- Email: [email protected]
DNA markers for the diagnosis of Wilson disease
Journal of Hepatology, 1993; 17:269-276
269
© 1993 ElsevierScientificPublishers Ireland Ltd. All rights reserved. 0168-8278/93/$06.00 HEPAT 01394
D N A markers for the diagnosis of Wilson disease R o d e r i c k H.J. H o u w e n a'b'l, Eve A. R o b e r t s "'b'°, G o r d o n R. T h o m a s "'c and D i a n e W. C o x a'b'c " The Research Institute. The Hospital for Sick Children. b Department of Pediatrics, "Department of Molecular and Medical Genetics, a Department of Medicine. University of Toronto. Toronto, Canada
(Received 11 May 1992)
Wilson disease is an autosomal recessive disorder of copper transport for which the basic defect is unknown. Laboratory diagnosis of Wilson disease is usually made by measuring serum ceruloplasmin concentration, urinary copper excretion, and liver copper concentration. However, discrimination between heterozygotes and patients is sometimes difficult. The gene for Wilson disease has been assigned to chromosome-13 at q14-q21. In this study, 10 markers from the 13q14-13q21 region were investigated in 12 families with a well-established diagnosis, to confirm reported linkage results. Markers from the same region were tested in two additional families, in which a sib of each index case had unclear results with conventional biochemical assays. The linkage results in this study are similar to those of Middle Eastern families, and support the hypothesis of a single disease locus. In the two families studied for diagnostic purposes, the status of 2 presymptomatic sibs was established as affected and 1 as unaffected. This study therefore shows that DNA markers can be used to discriminate between presymptomatic patients and non-affected individuals when biochemical results are equivocal, as long as an index case with Wilson disease of known status is available and markers are informative. K e y words." Wilson disease; Copper; Restriction fragment length polymorphisms; DNA diagnosis
Wilson disease (Wilson's disease*), an autosomal recessive disorder of copper transport, is characterized by accumulation of copper in the liver, and secondarily in the brain, cornea and kidneys (1). Hepatic accumulation of copper is caused by decreased biliary copper excretion, but the underlying biochemical defect is unknown. Treatment involves the removal of copper from tissues by the chelating agents penicillamine or trientine, although the use of zinc has recently been advocated (2,3). Patients presenting with the hepatic or neurological form of the disease are usually diagnosed by the presence of corneal copper deposits (Kayser-Fleischer rings), a low serum ceruloplasmin concentration (<180 mg/1) and elevated urinary copper excretion, both basal and after penicillamine administration (1,4,5). Elevated hepatic copper concentration is usually also a reliable diagnostic test for Wilson disease (6). However, apart from the technical problems of the assay, percutaneous liver
biopsy can be hazardous in patients with coagulopathy, or unduly invasive in sibs without definite clinical or biochemical signs of liver disease. The range of hepatic copper concentration in patients varies widely (6,7), and can overlap with that of heterozygotes (A. Sass-Kortsak, unpublished data). Patients with chronic cholestatic liver disease may have similar elevations of hepatic copper (6-9). Incorporation of a copper radioisotope into serum ceruloplasmin can be helpful in diagnosis, although not reliable in patients with a very low serum ceruloplasmin level (10). Thus diagnosis can be difficult to establish in some individuals. Distinguishing between patients and heterozygotes can be difficult, especially in a young person whose copper accumulation is too low to cause biochemical or clinical abnormalities, or in a patient started on chelation therapy before the diagnosis of Wilson disease has been clearly proven. Early diagnosis is important so that treatment can be initiated before tissue damage occurs
Correspondence to: Prof. Dr. D. W. Cox, The Hospital for Sick Children, 555 UniversityAvenue,Toronto, Ontario, Canada M5G IX8. Present address: Wilhelmina Children's Hospital, Utrecht, The Netherlands.
* Current nomenclature no longer includes's on disease names.
270
R.H.J. HOUWEN et al.
(11,12). Conversely, erroneous diagnosis of a heterozygote as a patient results in a lifetime of unnecessary medication, with possible serious adverse side effects. The gene for Wilson disease (WND) was assigned to chromosome-13 through linkage with the erythrocyte enzyme esterase D (13), and subsequently with other markers in the 13q14-13q21 region (14,15). These polymorphic D N A markers (restriction fragment length polymorphisms, RFLPs) have been used for carrier detection (16-18), and for prenatal diagnosis (18,19). We report the application of this approach to 3 individuals for which the diagnosis of Wilson disease, based on biochemical results, was ambiguous. We established a diagnosis of affected status in 2 cases, using D N A markers, and normal in a third. Our observations demonstrate the importance of this approach as an aid to the diagnosis of Wilson disease in sibs of known patients.
families had at least 2 available affected homozygotes, or 1 patient plus at least 3 unaffected sibs. All participants gave informed consent and the study protocol was approved by our local Ethics Committee. Eighteen individuals were affected and 33 sibs were normal. The diagnosis of Wilson disease was based on clinical symptoms, a serum ceruloplasmin below 180 mg/l, a urinary copper excretion above 1.6 l~mol/24 h, and a liver copper above 250/~g/g dry liver tissue, and in most cases the presence of Kayser-Fleischer rings. Clinical, genetic, and biochemical features have been reported previously (24). The markers used, listed in Table l, cover a total distance of approximately 45 centimorgans (cM), and flank the Wilson disease locus (18,25) (Fig. I). Our results using rodent x human hybrid cell lines are compatible with this marker order (26). Closest flanking markers are reported to be D13S31 and D13S59 (18). Genetic linkage was evaluated by pairwise testing of marker and disease loci using the computer program L I P E D (27).
Patients and Methods
Familiesfor diagnostic investigation Laboratory analysis D N A was prepared from EDTA-treated venous blood samples (20). Restriction enzyme digestion, agarose gel electrophoresis and Southern blot transfer were carried out as described (21). Hybridizations were carried out by the Church and Gilbert method (22). Autoradiography was performed at - 7 0 ° C for I - 7 days to reveal restriction fragment length polymorphisms.
Families usedfor linkage analysis A total of 12 Canadian two-generation families of Northern European ancestry (mainly British and French) was used to study the linkage between W N D and 10 markers at 9 loci in 13q14-13q21 (23) (Table 1). These
Parents and children of two families, recently referred to HSC for diagnostic investigation, were typed for the i0 polymorphic D N A markers used for the linkage study (Table I). In one family, the more distant markers D13S2 and D13S4 were used because closer markers were not informative. In another family, the additional marker D13S59, reported to be the closest distal flanking marker (18), was used.
Patients Family 1 (W46). The proband (II-l, Fig. 2) in this family presented at another hospital at the age of l0 years with acute liver failure and died within a few days. Wilson disease was diagnosed on the basis of the pres-
TABLE I Description of markers used in this study~ Locus D13S22 ESD RB 1 RB 1 DI3S31
Probe pG 14E3.8 pEL-22(3') p68RS2.0 p88R0.6 pCRI324
Map location 13ql4 13q14 13q 14 13q14 13q14
D 13S59b D 13S55 D13S26
pRI-4 CRI-R214 pH2-10
13q 14 13q 14-21 13q21
D 13S39 WC25 D13S41 WC83 D 13S12 pG 18E2.1 D13S2b p9Dll D13S4b pl E8 'Summarized from Ref.23. bNot tested in families used for linkage analysis.
13q21 13q21 13q21 13q22 13q22-31
Enzyme Hindlll Apal Rsal Xbal (I)Taql (2)PvulI Banll Taql (I)Hphl (2)BclI Mspl Taql Mspl Mspl Mspl
Alleles (kb) 2.3, 2.0 8.6, 7.0 8 alleles (2.0-1.5) 7.0, 5.5 6.7, 4.6 7.3, 5.5, 1.8 8.5, 4.9 8.9, 6.0 2.8, 2.0 6.3, 5.6 4.2, 2.4, 2.2 8.5, 2.4 8.8, 4.4 15.0, II.0, 10.5 10.1, 7.4
271
DNA DIAGNOSIS O F WILSON DISEASE
FAMILY 1 ( W 46) Male ( ) (cM)
Female (cM)
,
,C]'--
D13S22 6
6 ESD
3
3
3 0.1 0.1
1 2 7
11
5
4
13
6
5
3
3
2
4
0.1
6
RB1 D13S31 -91-Wilson disease D 13S59 gene D13S55 D13S26
1
1
1 1
2 2
1
4
ESD RB1 (p88) RB1(p68) D13S31
2 1 5 2
2
2
WND
N
N
W
D13S26 D13S2 D13S4 °++
2 2 1
2 1 2 2
2 2 1 2
D13S22
I I
'MI+
D 13S39 D13S41 D13S12 D 13S2 D13S4
I Fig. 1. Genetic map of the long arm of chromosome-13. Numbers represent approximate distances between loci, expressed as centimorgans. Values differ between males and females and are shown for both: values in males on the left, in females on the right, as calculated from 3-generation CEPH families (25). Position of the WND locus is from Ref. 18, in which estimated distances from DI3S31 and D13S59 are I and 3 centimorgans, respectively.
ence of Kayser-Fleischer rings, low serum ceruloplasmin concentration, increased serum copper concentration, and greatly increased urinary copper excretion (Table 2). Autopsy findings confirmed the diagnosis: the liver was cirrhotic with increased copper concentration demonstrated histochemically, and typical histopathological changes were present in the brain. At that time there were 2 sibs, aged 9 and 3 years. The 9-year-old brother (II-2) had hepatosplenomegaly but no Kayser-Fleischer rings. Serum ceruloplasmin concentration was low and basal urinary copper excretion was elevated (Table2). Liver biopsy showed steatosis, chronic active hepatitis and early cirrhosis, with a positive stain for parenchymal copper. Wilson disease was diagnosed and treatment with penicillamine was started. In contrast, the sister (II-3) had no hepatosplenomegaly. Her serum ceruloplasmin at age 3 was somewhat low at 120 mg/1, but urinary copper excretion was normal. Thus it was not clear whether she was a heterozygote or a patient, and she was re-assessed annually. There was no change clinically or biochemically until she was 7 years old. Although she had been apparently well, moderate hepatomegaly was noted, and spider angiomata were present. Serum aspartate aminotransferase (AST) was mildly elevated at 77 U]I (normal <36 U/l). Serum ceruloplasmin was low and urinary copper excretion was slightly elevated (Table 2). Kayser-
ESD RB1(p88) RB1(p68) D13S31
2 4 1
WND
D13S55 D13S26 D13S2 D13S4
2 2 2 2
I
I
2 2 4 2 W 2 2 1 2
2 2 4 2 W 2 2 1 2
21
I
5 2 N
'Hi
2 2 1
Fig. 2. DNA haplotypes in family (W 46). Only informative markers are shown. For each probe the longer fragment is designated 1, and the other 2. For the RBI probe, more alleles were recognized (Table 1). Inferred alleles for the Wilson disease locus (WND) are indicated by 'W' for the mutant allele and 'N' for the normal allele. The proband (II-I), indicated by an arrow-head, was deceased and no DNA was available. Solid symbol, affected homozygote; clear, unaffected; stippled, status unknown prior to DNA study. Chromosomes carrying normal alleles at the W N D locus are cross-hatched (paternal) or open (maternal). Chromosomes carrying Wilson disease alleles are solid (paternal) and stippled (maternal).
Fleischer rings were not detected by slit-lamp examination. Liver biopsy showed mild steatosis and scattered glycogenated liver cell nuclei on light microscopy. On electron microscopy, multiple features consistent with a diagnosis of Wilson disease were found, including enlarged and pleomorphic mitochondria with dilated tips of cristae, lipolysosomes, microvesicular steatosis, and abnormal bile canaliculi (28). On the basis of these studies, Wilson disease was diagnosed. DNA marker studies were initiated to verify this diagnosis. A fourth child (II-4), born after the proband died, was 3 years of age when his sister was diagnosed. His serum ceruloplasmin was normal (Table 2), but since infancy his liver had been palpable 2 cm below the right costal margin. Although thought to be probably unaffected, because of his equivocal sign, DNA marker studies were used to determine his genotype. Family 2 (W43). The proband in this family (II-2, Fig. 3), presented with depression, deteriorating school performance and a movement disorder at the age of 16. She was evaluated at another hospital and found to have Kayser-Fleischer rings, low serum ceruloplasmin
272
R.H.J. H O U W E N et al.
TABLE 2 Results of laboratory investigations in two families Family No.
Age= (yr)
Ceruloplasmin b (mg/l)
Serum copper b (pmol/l)
(#mol/24 h)
Urinary copper b
1 (W46)II-1 I1-2 II-3 II-4
10 9 7 3
low = I I0 181 281
51.9 5.4 10.0 ND d
26,9 12 0.7 ND
2 (W43)II-1 II-2 II-3
19 17 9
70 30 250
4.0 4.4 ND
2.0, 3.5, 1.6 4.8 ND
=Age at which stated results were obtained. bNormal values (Wilson disease values in parentheses): serum ceruloplasmin 180-450 mg/I (< 180 mg/I); serum copper 10-23/~mol/l (< 10 llmol/I); urinary copper <0.6/zmol/24 h (> 1.6 pmol/24 h). CAssayed at another hospital and numerical results could not be obtained. dND = not determined.
FAMILY 2 ( W 43 )
1
D13S22 ESD RBl(p88) D13S31
21I1
2 1 N 1
WND
D13S59 D13S55 D13S26 D13S39 D13S12
2 2 1
2 2
1 1
1 N
1 W
2 2 2 2
t 1 1 2
1
2
2
4 D13S22 ESD RB1 (p88) D13S31
1 2 2 2
1 1 1 1
WND
W
W
D13S59 D13S55 D13S26 D13S39 D13S12
2 2 1 1 1
1 1 1 2 2
R ~2
'11' wit w 1
1
2
1 1
2 2
1 2 "~'-R
1 1
2 1
1 1 2 1
2 2 2 1
1 1 2 2 1
2 2 2 1
Fig. 3. D N A haplotypes for family 2 (W 43). Symbols are as for Fig. 2. 'R' beside closed arrows indicates location of recombination events (see text). A recombination event has occurred on both paternal and maternal chromosomes in 1I-2. The double arrow indicates that the recombinant event in the paternal chromosome could be at either of the two indicated positions.
and elevated basal urinary copper excretion (Table 2). Wilson disease was diagnosed and penicillamine was started. As her neurological condition continued to worsen, she was referred to The Hospital for Sick Children for further treatment. On examination, in addition to neurological findings, hepatomegaly and abnormal liver function tests were found. A liver biopsy showed possible cirrhosis, with steatosis, cellular degenerative changes and minimal stainable copper. The proband's younger sister (1I-3) was well, normal
on physical examination, and had a normal serum ceruloplasmin concentration (Table 2). She was considered unaffected. However, the older brother (II-1), at 19 years of age, had low serum ceruloplasmin and mildly abnormal liver function tests (serum aspartate aminotransferase 81 U/I). Because he was eager to begin a job in a remote part of the country, he was started on penicillamine before the diagnostic evaluation was completed. It was later established that he did not have Kayser-Fleischer rings, as originally thought. Other causes of an elevated AST level had not been excluded. He was admitted to another hospital for further assessment. Urinary copper excretion was elevated (but he was still taking penicillamine); liver biopsy showed mild fibrosis but no features specifically diagnostic of Wilson disease (hepatic copper concentration was not measured). As he was unwilling to travel to Toronto, further studies were performed locally. Penicillamine was stopped for 6 months and urinary copper excretion repeatedly checked: without peniciUamine, copper excretion was slightly elevated (Table 2). Thus the diagnosis remained in doubt. The low serum ceruloplasmin was particularly difficult to evaluate because his mother (1-2) also had a low ceruloplasmin of 130 mg/l. DNA marker studies were therefore initiated to aid in diagnosis. The patient was placed on treatment with zinc sulphate pending the outcome of these studies.
Results
Linkage studies In these families of Northern European ancestry, several markers in the 13q 14 region (D 13S22, ESD, RB 1, D13S31) showed no recombination with WND by twopoint linkage analysis. By haplotyping, two recombinants were identified: one between ESD and RBI, and
DNA DIAGNOSIS OF WILSON DISEASE
273
one between RB1 and DI3S31. Markers D13S22, ESD and RBI had lod scores greater than 3, or odds of greater than 1000:1 against observed results being due to chance (Table3). Markers D13S31, D13S55 and D13S26 showed close linkage at lod scores of greater than 2. Distal markers D13S39, D13S41 and D13S12, showed high recombination (0) and insignificant lod scores because the limited number of meioses is inadequate to detect more distant linkage. These results support those from larger series of families from the Middle East and Europe (15-18,29). Our selection of markers for the diagnostic studies is therefore appropriate for this population.
rarely presents clinically before 5 years of age. With D N A - m a r k e r studies, he has now been diagnosed as homozygous normal. The summed chance of having an undetected double recombination including the Wilson disease gene in either of the parental chromosomes is approximately 2.3%. Therefore, the overall reliability of excluding the heterozygous state is greater than 97%. In the older brother of the proband in family 2 (W43), a provisional diagnosis of Wilson disease was made, mainly because of a low serum ceruloplasmin, which is known to be unreliable. Repeated assays of urinary copper excretion after penicillamine had been discontinued for 4 weeks to 6 months were all slightly elevated and the diagnosis remained in doubt (Table 2). The subject (II-1) and his affected sister (II-2) are genotypically non-identical, explainable by recombination events. A recombination event in the proband (11-2) on the maternal chromosome close to the W N D locus initially made DNA-based diagnosis in this family impossible. The W N D gene could be placed on either of the mother's chromosomes. However, with the use of the newer marker D13S59, which was informative in this family, the position of the recombinant event on the maternal chromosome could be identified as distal to the W N D locus. The brother is therefore affected and penicillamine therapy has been resumed. The chance ot" having a nondetected recombination event is less than 1%.
Diagnostic studies In family 1 (W46), the diagnosis of Wilson disease in the brother (Fig. 2, I1-2) was made on the basis of a low serum ceruloplasmin and copper level, increased urinary copper excretion (Table 2), as well as clinical and pathological signs (28). In the sister (II-3), the diagnosis might be questioned because clinical findings were subtle and the urinary copper excretion was only marginally increased, while low ceruloplasmin, levels could be due to the heterozygous state. However, her diagnosis of Wilson disease was confirmed using D N A markers because she and her affected brother were genotypically identical for the total chromosomal region investigated (Fig. 2). Nine markers were heterozygous in one or both parents and therefore informative for recombination. On the basis of reported recombination values from large C E P H families (25), the chance of having an undetected double recombination flanking the disease gene in 1I-3 is less than 0.01% for the paternal chromosome, and 2.3% for the maternal chromosome. Thus, the overall reliability of the diagnosis in this case is greater than 97%. There were no biochemical indications that the younger brother (11-4) was affected, but Wilson disease
Discussion Distinguishing patients with Wilson disease from heterozygotes is sometimes extremely difficult. Heterozygotes may show some of the biochemical characteristics of patients, particularly a low serum ceruloplasmin concentration, which is characteristic of heterozygotes in some families previously designated as 'atypical' (24). An
TABLE 3 Two-point linkage with the Wilson disease locus (WND) Locus
Recombination (01
0.00 0.001 0.01 0.05 D 13S22 3.25 3.24 3.16 2.84 ESD 3.38 3.38 3.34 3.09 RB1 6.38 6.36 6.20 5.50 D 13S31 2.76 2.75 2.67 2.29 D 13S55 - oo 1.14 2.04 2.29 D 13S26 - oo 0.69 1.64 2.09 D 13S39 - oo - 0.67 0.30 0.82 D 13S41 - oo - 1.60 - 0.62 0.00 DI3SI2 -oo -1.85 -0.85 -0.21 Linkage results are based on 12 caucasian Canadian families.
0. I 2.42 2.70 4.63 1.82 2.04 1.98 0.89 0.17 -0.01
0.2 1.59 1.83 2.95 0.98 1.30 1.35 0.68 0.21 0.13
0.3 0.82 0.98 1.47 0.40 0.62 0.70 0.36 0.13 0.10
0.4 0.23 0.29 0.40 0.09 0.16 0.20 0.10 0.04 0.03
Zm,x 3.25 3.38 6.38 2.76 2.31 2.09 0.89 0.23 0.14
0=,, 0.0 0.0 0.0 0.0 0.04 0.06 0.09 0.17 0.21
274 isolated low serum ceruloplasmin level therefore cannot be interpreted as diagnostic for Wilson disease (1,4,24), as seen in family 2 in which a parent has a low serum ceruloplasmin level. It is recommended that the diagnosis of asymptomatic patients be based on a combination of biochemical tests, including a low serum ceruloplasmin level, increased urinary copper excretion (basal and after penicillamine challenge) and increased hepatic copper concentration on liver biopsy (1,4,5,11). This latter procedure is invasive, carries a small risk, and can give unreliable results due to sample contamination or variations within different parts of the liver. For these reasons it cannot be justified in the absence of signs or symptoms. Likewise, the administration of copper radioisotopes to study copper incorporation into ceruloplasmin (absent in Wilson disease) also poses potential risks and can be unreliable in individuals with low serum ceruloplasmin concentrations (10,24). This was shown in 4 individuals in a family described previously, all of whom had low ceruloplasmin concentrations, lack of detectable incorporation of copper, and an elevation of liver copper. These characteristics initially led to the diagnosis of affected for two heterozygotes, who remained clinically asymptomatic for 20 years with no therapy (30). In initial linkage studies using esterase-D electrophoretic variants only, these two heterozygotes were still considered as patients with a recombination event (31). Copper radioisotope studies should therefore be used with caution in making a diagnosis. Thus the present diagnostic methods involve some risk, and may still be inconclusive in certain families. The development of DNA diagnosis of Wilson disease is of considerable importance. Diagnosis using DNA markers potentially provides a clear distinction between patients and heterozygotes in families with Wilson disease, as illustrated in the families presented here. In family 1, the diagnosis of 'affected' for the sister could have been established years earlier if DNA markers had been available, and before the onset of any clinical signs. If normal, extensive follow-up would not have been necessary. In a second sib the diagnosis of 'normal' was made in the presence of an equivocal sign - - infant hepatomegaly. In family 2, treatment was initiated before the diagnosis was certain and then diagnosis was difficult by the usual biochemical criteria. DNA studies confirmed the affected status. The devastating hepatic and neurological symptoms of Wilson disease can be prevented by life-long treatment with copper-chelating agents, particulary penicillamine, which are effective in achieving a negative total body copper balance. When treatment is given to presymptomatic patients, tissue damage is prevented and the overall
R.H.J. H O U W E N et al.
course is favorable (11,12). An accurate diagnosis for presymptomatic siblings of an index case is therefore essential. If a heterozygote or normal unaffected individual is treated with penicillamine, serious side-effects may result. In patients who do not have Wilson disease, e.g. rheumatoid arthritis or biliary cirrhosis, serious sideeffects occur in about 33% (32). Even for patients with Wilson disease, up to 20% experience some side-effects, half of which are serious, such as systemic lupus erythematosus or immune-complex nephritis (33). Therefore penicillamine should not be given to unaffected individuals. The use of DNA-based diagnosis requires an index case with an unequivocal diagnosis of Wilson disease, until the gene and its mutation are identified. This patient's chromosome markers serve as a reference for identification of the disease-carrying chromosomes. When the chromosomes of the sibs are studied, the chromosomes inherited from each parent can be identified, indicating whether or not the disease gene has been inherited. The sibs can be classified as patients, heterozygotes or homozygous normals. Furthermore, as our results show, with this approach 100% reliability cannot be obtained, because the diagnostic method is based on the use of flanking markers and there is always a chance of an undetected double recombinant, especially when only a few loci in the region of interest are informative in a particular individual. However, the chance of having a double recombinant can be calculated. Reliability will exceed 95% in most cases, because a large and growing number of markers on both sides of the Wilson disease gene are available. Another potential source of error in the diagnosis of Wilson disease by linked markers is the existence of more than one gene causing Wilson disease. There is some clinical evidence pointing to genetic heterogeneity, especially the observation that the affected members of a given family display similar clinical and biochemical phenotypes, while there is considerable phenotypic heterogeneity between pedigrees (24). This variation could be a result of different mutations in the same gene, from mutations at different loci, or from the action of environmental factors such as copper intake (or a combination of the above). If there were more than one locus for Wilson disease, DNA diagnosis would be impossible at present. However, current data give no support for more than one W N D locus. The gene is apparently localized on the same part of chromosome-13 in all populations studied to date: Middle-Eastern (13-15,18), Northern European (29,34), Italian (17), Chinese (35) white American (18,36) and the Canadian (Northern European ancestry) families in this report. Two studies
275
DNA DIAGNOSIS OF WILSON DISEASE
suggesting looser linkage to ESD were due to misclassification of heterozygotes as patients, as described above (31), or to broad confidence intervals associated with small sample size (29,31,34). Ultimately, the study of markers in the vicinity of the Wilson disease gene will lead to the isolation of the gene itself. This should lead to a real understanding of the pathogenesis of Wilson disease based on the gene function. It will also define the basis for heterogeneity in the clinical presentation and course of the disease. When the gene is isolated, the diagnosis of Wilson disease will also be possible by DNA methods in patients with liver or neurological diseases without a family history of this disease.
Acknowledgements We thank Prasanna Kumaranayake for technical assistance, Gail Billingsley for data verification, Drs. A. Goodridge, F.H. Anderson and W. Mackle for assistance in patient follow-up, and Drs. A.E. Bale, C.H.C.M. Buys, W. Cavenee, T. Dryja and J. Squire for DNA probes. This study was supported by the Physicians' Services Incorporated Foundation, R.H.J.H. was supported by a fellowship from the Ter Meulen Fund of the Royal Netherlands Academy of Arts and Sciences.
References I Danks DM. Disorders of copper transport. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic Basis of Inherited Disease. New York: McGraw-Hill, 1989:1411-31. 2 Hill GM, Brewer G J, Prasad AS, Hydrick CR, Hartmann DE. Treatment of Wilson's disease with zinc. I. Oral zinc therapy regimens. Hepatology 1987; 7: 522-8. 3 Hoogenraad TU, van Hattum J, van den Hamer CJA. Management of Wilson's disease with zinc sulphate. Experience in a series of 27 patients. J Neurol Sci 1987; 77: 137-46. 4 Scheinberg IH, Sternlieb I. Wilson's Disease. Philadelphia: WB Saunders, 1984. 5 Martins da Costa C, Baldwin D, Portmann B, Lolin Y, Mowat AP, Mieli-Vergani G. Value of urinary copper excretion after penicillamine challenge in the diagnosis of Wilson's disease. Hepatology 1992; 15: 609-15. 6 Evans J, Newman S, Sherlock S. Liver copper levels in intrahepatlc cholestasis of childhood. Gastroenterology 1978; 75: 875-8. 7 Smallwood RA, Wdliams HA, Rosenoer VM, Sherlock S. Lwer copper levels in liver disease: studies using neutron activation analysis. Lancet 1968; ii: 1310-3. 8 Danks DM. Copper and liver disease. Eur J Pediatr 1991; 150: 142-8. 9 Bhave SA, Pandit AN, Pradhan AM, et al. Liver disease in India. Arch Dis Child 1982: 57: 922-8. 10 Sternlieb I, Scheinberg I H. The role of radiocopper in the diagnosis of Wilson's disease. Gastroenterology 1979; 77: 138-42. I1 Sternlieb I, Scheinberg IH. Prevention of Wilson's disease in asymptomatic patients. N Engl J Med 1968; 278: 352-9. 12 Walshe JM. Diagnosis and treatment of pre-symptomatic Wilson's disease. Lancet 1988; ii: 435-7. 13 Frydman M, Bonne-Tamir B, Farrer LA, et al. Assignment of the gene for Wilson disease to chromosome 13: linkage to the esterase D locus. Proc Natl Acad Sci USA 1985; 82: 1819-21. 14 Bowcock AM, Farrer LA, Cavalh-Sforza LL et al. Mapping the Wilson disease locus to a cluster of linked polymorphic markers on chromosome 13. Am J Hum Genet 1987; 41: 27-35. 15 Bowcock AM, Farrer LA, Hebert JM, et al. Eight closely linked loci place the Wilson disease locus within 13q14-q21. Am J Hum Genet 1988; 43: 664-74. 16 Farter LA, Bonne-Tamir B, Frydman M, et al. Predicting genotypes at loci for autosomal recessive disorders using linked genetic markers: applicatton to Wilson disease. Hum Genet 1988; 79: 109-17. 17 Figus A, Lampis R, Devoto M, et al. Carrier detection and early
18
19
20
21
22 23
24
25
26
27
28
29
30
31
diagnosis of Wilson's disease by restriction fragment length polymorphism analysis. J Med Genet 1989; 26: 78-82. Farrer LA, Bowcock AM, Hebert JM et al. Predictive testing for Wilson disease using tightly linked and flanking markers. Neurology 1991; 41: 992-9. Cossu P, Pirastu M, Nucaro A, et al. Prenatal diagnosis of Wilson's disease by analysis of DNA polymorphisms. N Engl J Med 1992; 327: 57. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988; 16: 1215. Benger JC, Cox DW. Polymorphisms of the imrrmnoglobulin heavy chain delta gene and association with other constant region genes. Am J Hum Genet 1989; 45: 606-14. Church GM, Gilbert W. Genomic sequencing. Proc Natl Acad Sci USA 1984; 81: 1991-5. Williamson R, Bowcock A, Kidd K, et al. Report of the DNA committee and catalogues of cloned and mapped genes, markers formatted for PCR and DNA polymorphisms. HGM Workshop I 1. Cytogenet Cell Genet 1991; 58:1190-1832. Cox DW, Fraser FC, Sass-Kortsak A. A genetic study of Wilson's disease: evidence for heterogeneity. Am J Hum Genet 1972; 24: 646-66. Bowcock AM, Farrer LA, Hebert JM, Bale AE, Cavalli-Sforza L. A contiguous linkage map of chromosome 13q with 39 distinct loci separated on average by 5.1 centimorgans. Genomics 1991; I 1: 517-29. Houwen RHJ, Pautler SE, Barwell JA, et al. Isolation and regional localization of 25 anonymous DNA probes on a chromosome 13 hybrid panel. Cytogenet Cell Genet 1991; 57: 87-90. Ott J. Estimation of the recombination fraction in human pedigrees: efficient computation of the likelihood for human linkage studies. Am J Hum Genet 1974; 26: 588-97. Phillips M J, Poucell S, Patterson J, Valencia P. The Liver. An Atlas and Text of Ultrastructural Pathology. New York: Raven Press, 1987. Houwen RHJ, Scheffer H, te Meerman G J, van der Vlies P, Buys CHCM. Close linkage of the Wilsons disease locus to D13S12 in the chromosomal region 13q21 and not to ESD in 13q14. Hum Genet 1990; 85: 560-2. Cox DW, Billingsley GD. The application of DNA markers to the diagnosis of presymptomatic Wilson disease. In: Wetterberg L, ed. Proceedings of Genetics of Psychiatric Diseases. Wenner-Gren International Symposium, Vol. 51. Stockholm: Macmillan Press, 1988: 167-77. Cox DW, Weitkamp LR, Lewis M, Sass-Kortsak A. Linkage
276 studies in Wilson's disease (hepatolenticular degeneration). Cytogenet Cell Genet 1985; 40: 608-9. 32 James OFW. D-Penicillamine for primary biliary cirrhosis. Gut 1985; 26: 109-13. 33 Walshe JM. Wilson's disease presenting with features of hepatic dysfunction: a clinical analysis of eighty-seven patients. Q J Med New Ser 1989; 70: 253-63. 34 Scheffer H, Houwen RHJ, te Meerman GJ et al. Identification of
R.H.J. HOUWEN et al. crossovers in Wilson disease families as reference points for a genetic localization of the gene. Hum Genet 1992; 89: 607-I I. 35 Chuang L-M, Tai T-Y, Wang T-R et al. Esterase D and retinoblastoma gene loci are tightly linked to Wilson's disease in Chinese pedigrees from Taiwan. Hum Genet 1991; 87: 465-8. 36 Yuzbasiyan-Gurkan V, Brewer G J, Boerwinkle E, Venta PJ. Linkage of the Wilson disease gene to chromosome 13 in NorthAmerican pedigrees. Am J Hum Genet 1988; 42: 825-9.
269
© 1993 ElsevierScientificPublishers Ireland Ltd. All rights reserved. 0168-8278/93/$06.00 HEPAT 01394
D N A markers for the diagnosis of Wilson disease R o d e r i c k H.J. H o u w e n a'b'l, Eve A. R o b e r t s "'b'°, G o r d o n R. T h o m a s "'c and D i a n e W. C o x a'b'c " The Research Institute. The Hospital for Sick Children. b Department of Pediatrics, "Department of Molecular and Medical Genetics, a Department of Medicine. University of Toronto. Toronto, Canada
(Received 11 May 1992)
Wilson disease is an autosomal recessive disorder of copper transport for which the basic defect is unknown. Laboratory diagnosis of Wilson disease is usually made by measuring serum ceruloplasmin concentration, urinary copper excretion, and liver copper concentration. However, discrimination between heterozygotes and patients is sometimes difficult. The gene for Wilson disease has been assigned to chromosome-13 at q14-q21. In this study, 10 markers from the 13q14-13q21 region were investigated in 12 families with a well-established diagnosis, to confirm reported linkage results. Markers from the same region were tested in two additional families, in which a sib of each index case had unclear results with conventional biochemical assays. The linkage results in this study are similar to those of Middle Eastern families, and support the hypothesis of a single disease locus. In the two families studied for diagnostic purposes, the status of 2 presymptomatic sibs was established as affected and 1 as unaffected. This study therefore shows that DNA markers can be used to discriminate between presymptomatic patients and non-affected individuals when biochemical results are equivocal, as long as an index case with Wilson disease of known status is available and markers are informative. K e y words." Wilson disease; Copper; Restriction fragment length polymorphisms; DNA diagnosis
Wilson disease (Wilson's disease*), an autosomal recessive disorder of copper transport, is characterized by accumulation of copper in the liver, and secondarily in the brain, cornea and kidneys (1). Hepatic accumulation of copper is caused by decreased biliary copper excretion, but the underlying biochemical defect is unknown. Treatment involves the removal of copper from tissues by the chelating agents penicillamine or trientine, although the use of zinc has recently been advocated (2,3). Patients presenting with the hepatic or neurological form of the disease are usually diagnosed by the presence of corneal copper deposits (Kayser-Fleischer rings), a low serum ceruloplasmin concentration (<180 mg/1) and elevated urinary copper excretion, both basal and after penicillamine administration (1,4,5). Elevated hepatic copper concentration is usually also a reliable diagnostic test for Wilson disease (6). However, apart from the technical problems of the assay, percutaneous liver
biopsy can be hazardous in patients with coagulopathy, or unduly invasive in sibs without definite clinical or biochemical signs of liver disease. The range of hepatic copper concentration in patients varies widely (6,7), and can overlap with that of heterozygotes (A. Sass-Kortsak, unpublished data). Patients with chronic cholestatic liver disease may have similar elevations of hepatic copper (6-9). Incorporation of a copper radioisotope into serum ceruloplasmin can be helpful in diagnosis, although not reliable in patients with a very low serum ceruloplasmin level (10). Thus diagnosis can be difficult to establish in some individuals. Distinguishing between patients and heterozygotes can be difficult, especially in a young person whose copper accumulation is too low to cause biochemical or clinical abnormalities, or in a patient started on chelation therapy before the diagnosis of Wilson disease has been clearly proven. Early diagnosis is important so that treatment can be initiated before tissue damage occurs
Correspondence to: Prof. Dr. D. W. Cox, The Hospital for Sick Children, 555 UniversityAvenue,Toronto, Ontario, Canada M5G IX8. Present address: Wilhelmina Children's Hospital, Utrecht, The Netherlands.
* Current nomenclature no longer includes's on disease names.
270
R.H.J. HOUWEN et al.
(11,12). Conversely, erroneous diagnosis of a heterozygote as a patient results in a lifetime of unnecessary medication, with possible serious adverse side effects. The gene for Wilson disease (WND) was assigned to chromosome-13 through linkage with the erythrocyte enzyme esterase D (13), and subsequently with other markers in the 13q14-13q21 region (14,15). These polymorphic D N A markers (restriction fragment length polymorphisms, RFLPs) have been used for carrier detection (16-18), and for prenatal diagnosis (18,19). We report the application of this approach to 3 individuals for which the diagnosis of Wilson disease, based on biochemical results, was ambiguous. We established a diagnosis of affected status in 2 cases, using D N A markers, and normal in a third. Our observations demonstrate the importance of this approach as an aid to the diagnosis of Wilson disease in sibs of known patients.
families had at least 2 available affected homozygotes, or 1 patient plus at least 3 unaffected sibs. All participants gave informed consent and the study protocol was approved by our local Ethics Committee. Eighteen individuals were affected and 33 sibs were normal. The diagnosis of Wilson disease was based on clinical symptoms, a serum ceruloplasmin below 180 mg/l, a urinary copper excretion above 1.6 l~mol/24 h, and a liver copper above 250/~g/g dry liver tissue, and in most cases the presence of Kayser-Fleischer rings. Clinical, genetic, and biochemical features have been reported previously (24). The markers used, listed in Table l, cover a total distance of approximately 45 centimorgans (cM), and flank the Wilson disease locus (18,25) (Fig. I). Our results using rodent x human hybrid cell lines are compatible with this marker order (26). Closest flanking markers are reported to be D13S31 and D13S59 (18). Genetic linkage was evaluated by pairwise testing of marker and disease loci using the computer program L I P E D (27).
Patients and Methods
Familiesfor diagnostic investigation Laboratory analysis D N A was prepared from EDTA-treated venous blood samples (20). Restriction enzyme digestion, agarose gel electrophoresis and Southern blot transfer were carried out as described (21). Hybridizations were carried out by the Church and Gilbert method (22). Autoradiography was performed at - 7 0 ° C for I - 7 days to reveal restriction fragment length polymorphisms.
Families usedfor linkage analysis A total of 12 Canadian two-generation families of Northern European ancestry (mainly British and French) was used to study the linkage between W N D and 10 markers at 9 loci in 13q14-13q21 (23) (Table 1). These
Parents and children of two families, recently referred to HSC for diagnostic investigation, were typed for the i0 polymorphic D N A markers used for the linkage study (Table I). In one family, the more distant markers D13S2 and D13S4 were used because closer markers were not informative. In another family, the additional marker D13S59, reported to be the closest distal flanking marker (18), was used.
Patients Family 1 (W46). The proband (II-l, Fig. 2) in this family presented at another hospital at the age of l0 years with acute liver failure and died within a few days. Wilson disease was diagnosed on the basis of the pres-
TABLE I Description of markers used in this study~ Locus D13S22 ESD RB 1 RB 1 DI3S31
Probe pG 14E3.8 pEL-22(3') p68RS2.0 p88R0.6 pCRI324
Map location 13ql4 13q14 13q 14 13q14 13q14
D 13S59b D 13S55 D13S26
pRI-4 CRI-R214 pH2-10
13q 14 13q 14-21 13q21
D 13S39 WC25 D13S41 WC83 D 13S12 pG 18E2.1 D13S2b p9Dll D13S4b pl E8 'Summarized from Ref.23. bNot tested in families used for linkage analysis.
13q21 13q21 13q21 13q22 13q22-31
Enzyme Hindlll Apal Rsal Xbal (I)Taql (2)PvulI Banll Taql (I)Hphl (2)BclI Mspl Taql Mspl Mspl Mspl
Alleles (kb) 2.3, 2.0 8.6, 7.0 8 alleles (2.0-1.5) 7.0, 5.5 6.7, 4.6 7.3, 5.5, 1.8 8.5, 4.9 8.9, 6.0 2.8, 2.0 6.3, 5.6 4.2, 2.4, 2.2 8.5, 2.4 8.8, 4.4 15.0, II.0, 10.5 10.1, 7.4
271
DNA DIAGNOSIS O F WILSON DISEASE
FAMILY 1 ( W 46) Male ( ) (cM)
Female (cM)
,
,C]'--
D13S22 6
6 ESD
3
3
3 0.1 0.1
1 2 7
11
5
4
13
6
5
3
3
2
4
0.1
6
RB1 D13S31 -91-Wilson disease D 13S59 gene D13S55 D13S26
1
1
1 1
2 2
1
4
ESD RB1 (p88) RB1(p68) D13S31
2 1 5 2
2
2
WND
N
N
W
D13S26 D13S2 D13S4 °++
2 2 1
2 1 2 2
2 2 1 2
D13S22
I I
'MI+
D 13S39 D13S41 D13S12 D 13S2 D13S4
I Fig. 1. Genetic map of the long arm of chromosome-13. Numbers represent approximate distances between loci, expressed as centimorgans. Values differ between males and females and are shown for both: values in males on the left, in females on the right, as calculated from 3-generation CEPH families (25). Position of the WND locus is from Ref. 18, in which estimated distances from DI3S31 and D13S59 are I and 3 centimorgans, respectively.
ence of Kayser-Fleischer rings, low serum ceruloplasmin concentration, increased serum copper concentration, and greatly increased urinary copper excretion (Table 2). Autopsy findings confirmed the diagnosis: the liver was cirrhotic with increased copper concentration demonstrated histochemically, and typical histopathological changes were present in the brain. At that time there were 2 sibs, aged 9 and 3 years. The 9-year-old brother (II-2) had hepatosplenomegaly but no Kayser-Fleischer rings. Serum ceruloplasmin concentration was low and basal urinary copper excretion was elevated (Table2). Liver biopsy showed steatosis, chronic active hepatitis and early cirrhosis, with a positive stain for parenchymal copper. Wilson disease was diagnosed and treatment with penicillamine was started. In contrast, the sister (II-3) had no hepatosplenomegaly. Her serum ceruloplasmin at age 3 was somewhat low at 120 mg/1, but urinary copper excretion was normal. Thus it was not clear whether she was a heterozygote or a patient, and she was re-assessed annually. There was no change clinically or biochemically until she was 7 years old. Although she had been apparently well, moderate hepatomegaly was noted, and spider angiomata were present. Serum aspartate aminotransferase (AST) was mildly elevated at 77 U]I (normal <36 U/l). Serum ceruloplasmin was low and urinary copper excretion was slightly elevated (Table 2). Kayser-
ESD RB1(p88) RB1(p68) D13S31
2 4 1
WND
D13S55 D13S26 D13S2 D13S4
2 2 2 2
I
I
2 2 4 2 W 2 2 1 2
2 2 4 2 W 2 2 1 2
21
I
5 2 N
'Hi
2 2 1
Fig. 2. DNA haplotypes in family (W 46). Only informative markers are shown. For each probe the longer fragment is designated 1, and the other 2. For the RBI probe, more alleles were recognized (Table 1). Inferred alleles for the Wilson disease locus (WND) are indicated by 'W' for the mutant allele and 'N' for the normal allele. The proband (II-I), indicated by an arrow-head, was deceased and no DNA was available. Solid symbol, affected homozygote; clear, unaffected; stippled, status unknown prior to DNA study. Chromosomes carrying normal alleles at the W N D locus are cross-hatched (paternal) or open (maternal). Chromosomes carrying Wilson disease alleles are solid (paternal) and stippled (maternal).
Fleischer rings were not detected by slit-lamp examination. Liver biopsy showed mild steatosis and scattered glycogenated liver cell nuclei on light microscopy. On electron microscopy, multiple features consistent with a diagnosis of Wilson disease were found, including enlarged and pleomorphic mitochondria with dilated tips of cristae, lipolysosomes, microvesicular steatosis, and abnormal bile canaliculi (28). On the basis of these studies, Wilson disease was diagnosed. DNA marker studies were initiated to verify this diagnosis. A fourth child (II-4), born after the proband died, was 3 years of age when his sister was diagnosed. His serum ceruloplasmin was normal (Table 2), but since infancy his liver had been palpable 2 cm below the right costal margin. Although thought to be probably unaffected, because of his equivocal sign, DNA marker studies were used to determine his genotype. Family 2 (W43). The proband in this family (II-2, Fig. 3), presented with depression, deteriorating school performance and a movement disorder at the age of 16. She was evaluated at another hospital and found to have Kayser-Fleischer rings, low serum ceruloplasmin
272
R.H.J. H O U W E N et al.
TABLE 2 Results of laboratory investigations in two families Family No.
Age= (yr)
Ceruloplasmin b (mg/l)
Serum copper b (pmol/l)
(#mol/24 h)
Urinary copper b
1 (W46)II-1 I1-2 II-3 II-4
10 9 7 3
low = I I0 181 281
51.9 5.4 10.0 ND d
26,9 12 0.7 ND
2 (W43)II-1 II-2 II-3
19 17 9
70 30 250
4.0 4.4 ND
2.0, 3.5, 1.6 4.8 ND
=Age at which stated results were obtained. bNormal values (Wilson disease values in parentheses): serum ceruloplasmin 180-450 mg/I (< 180 mg/I); serum copper 10-23/~mol/l (< 10 llmol/I); urinary copper <0.6/zmol/24 h (> 1.6 pmol/24 h). CAssayed at another hospital and numerical results could not be obtained. dND = not determined.
FAMILY 2 ( W 43 )
1
D13S22 ESD RBl(p88) D13S31
21I1
2 1 N 1
WND
D13S59 D13S55 D13S26 D13S39 D13S12
2 2 1
2 2
1 1
1 N
1 W
2 2 2 2
t 1 1 2
1
2
2
4 D13S22 ESD RB1 (p88) D13S31
1 2 2 2
1 1 1 1
WND
W
W
D13S59 D13S55 D13S26 D13S39 D13S12
2 2 1 1 1
1 1 1 2 2
R ~2
'11' wit w 1
1
2
1 1
2 2
1 2 "~'-R
1 1
2 1
1 1 2 1
2 2 2 1
1 1 2 2 1
2 2 2 1
Fig. 3. D N A haplotypes for family 2 (W 43). Symbols are as for Fig. 2. 'R' beside closed arrows indicates location of recombination events (see text). A recombination event has occurred on both paternal and maternal chromosomes in 1I-2. The double arrow indicates that the recombinant event in the paternal chromosome could be at either of the two indicated positions.
and elevated basal urinary copper excretion (Table 2). Wilson disease was diagnosed and penicillamine was started. As her neurological condition continued to worsen, she was referred to The Hospital for Sick Children for further treatment. On examination, in addition to neurological findings, hepatomegaly and abnormal liver function tests were found. A liver biopsy showed possible cirrhosis, with steatosis, cellular degenerative changes and minimal stainable copper. The proband's younger sister (1I-3) was well, normal
on physical examination, and had a normal serum ceruloplasmin concentration (Table 2). She was considered unaffected. However, the older brother (II-1), at 19 years of age, had low serum ceruloplasmin and mildly abnormal liver function tests (serum aspartate aminotransferase 81 U/I). Because he was eager to begin a job in a remote part of the country, he was started on penicillamine before the diagnostic evaluation was completed. It was later established that he did not have Kayser-Fleischer rings, as originally thought. Other causes of an elevated AST level had not been excluded. He was admitted to another hospital for further assessment. Urinary copper excretion was elevated (but he was still taking penicillamine); liver biopsy showed mild fibrosis but no features specifically diagnostic of Wilson disease (hepatic copper concentration was not measured). As he was unwilling to travel to Toronto, further studies were performed locally. Penicillamine was stopped for 6 months and urinary copper excretion repeatedly checked: without peniciUamine, copper excretion was slightly elevated (Table 2). Thus the diagnosis remained in doubt. The low serum ceruloplasmin was particularly difficult to evaluate because his mother (1-2) also had a low ceruloplasmin of 130 mg/l. DNA marker studies were therefore initiated to aid in diagnosis. The patient was placed on treatment with zinc sulphate pending the outcome of these studies.
Results
Linkage studies In these families of Northern European ancestry, several markers in the 13q 14 region (D 13S22, ESD, RB 1, D13S31) showed no recombination with WND by twopoint linkage analysis. By haplotyping, two recombinants were identified: one between ESD and RBI, and
DNA DIAGNOSIS OF WILSON DISEASE
273
one between RB1 and DI3S31. Markers D13S22, ESD and RBI had lod scores greater than 3, or odds of greater than 1000:1 against observed results being due to chance (Table3). Markers D13S31, D13S55 and D13S26 showed close linkage at lod scores of greater than 2. Distal markers D13S39, D13S41 and D13S12, showed high recombination (0) and insignificant lod scores because the limited number of meioses is inadequate to detect more distant linkage. These results support those from larger series of families from the Middle East and Europe (15-18,29). Our selection of markers for the diagnostic studies is therefore appropriate for this population.
rarely presents clinically before 5 years of age. With D N A - m a r k e r studies, he has now been diagnosed as homozygous normal. The summed chance of having an undetected double recombination including the Wilson disease gene in either of the parental chromosomes is approximately 2.3%. Therefore, the overall reliability of excluding the heterozygous state is greater than 97%. In the older brother of the proband in family 2 (W43), a provisional diagnosis of Wilson disease was made, mainly because of a low serum ceruloplasmin, which is known to be unreliable. Repeated assays of urinary copper excretion after penicillamine had been discontinued for 4 weeks to 6 months were all slightly elevated and the diagnosis remained in doubt (Table 2). The subject (II-1) and his affected sister (II-2) are genotypically non-identical, explainable by recombination events. A recombination event in the proband (11-2) on the maternal chromosome close to the W N D locus initially made DNA-based diagnosis in this family impossible. The W N D gene could be placed on either of the mother's chromosomes. However, with the use of the newer marker D13S59, which was informative in this family, the position of the recombinant event on the maternal chromosome could be identified as distal to the W N D locus. The brother is therefore affected and penicillamine therapy has been resumed. The chance ot" having a nondetected recombination event is less than 1%.
Diagnostic studies In family 1 (W46), the diagnosis of Wilson disease in the brother (Fig. 2, I1-2) was made on the basis of a low serum ceruloplasmin and copper level, increased urinary copper excretion (Table 2), as well as clinical and pathological signs (28). In the sister (II-3), the diagnosis might be questioned because clinical findings were subtle and the urinary copper excretion was only marginally increased, while low ceruloplasmin, levels could be due to the heterozygous state. However, her diagnosis of Wilson disease was confirmed using D N A markers because she and her affected brother were genotypically identical for the total chromosomal region investigated (Fig. 2). Nine markers were heterozygous in one or both parents and therefore informative for recombination. On the basis of reported recombination values from large C E P H families (25), the chance of having an undetected double recombination flanking the disease gene in 1I-3 is less than 0.01% for the paternal chromosome, and 2.3% for the maternal chromosome. Thus, the overall reliability of the diagnosis in this case is greater than 97%. There were no biochemical indications that the younger brother (11-4) was affected, but Wilson disease
Discussion Distinguishing patients with Wilson disease from heterozygotes is sometimes extremely difficult. Heterozygotes may show some of the biochemical characteristics of patients, particularly a low serum ceruloplasmin concentration, which is characteristic of heterozygotes in some families previously designated as 'atypical' (24). An
TABLE 3 Two-point linkage with the Wilson disease locus (WND) Locus
Recombination (01
0.00 0.001 0.01 0.05 D 13S22 3.25 3.24 3.16 2.84 ESD 3.38 3.38 3.34 3.09 RB1 6.38 6.36 6.20 5.50 D 13S31 2.76 2.75 2.67 2.29 D 13S55 - oo 1.14 2.04 2.29 D 13S26 - oo 0.69 1.64 2.09 D 13S39 - oo - 0.67 0.30 0.82 D 13S41 - oo - 1.60 - 0.62 0.00 DI3SI2 -oo -1.85 -0.85 -0.21 Linkage results are based on 12 caucasian Canadian families.
0. I 2.42 2.70 4.63 1.82 2.04 1.98 0.89 0.17 -0.01
0.2 1.59 1.83 2.95 0.98 1.30 1.35 0.68 0.21 0.13
0.3 0.82 0.98 1.47 0.40 0.62 0.70 0.36 0.13 0.10
0.4 0.23 0.29 0.40 0.09 0.16 0.20 0.10 0.04 0.03
Zm,x 3.25 3.38 6.38 2.76 2.31 2.09 0.89 0.23 0.14
0=,, 0.0 0.0 0.0 0.0 0.04 0.06 0.09 0.17 0.21
274 isolated low serum ceruloplasmin level therefore cannot be interpreted as diagnostic for Wilson disease (1,4,24), as seen in family 2 in which a parent has a low serum ceruloplasmin level. It is recommended that the diagnosis of asymptomatic patients be based on a combination of biochemical tests, including a low serum ceruloplasmin level, increased urinary copper excretion (basal and after penicillamine challenge) and increased hepatic copper concentration on liver biopsy (1,4,5,11). This latter procedure is invasive, carries a small risk, and can give unreliable results due to sample contamination or variations within different parts of the liver. For these reasons it cannot be justified in the absence of signs or symptoms. Likewise, the administration of copper radioisotopes to study copper incorporation into ceruloplasmin (absent in Wilson disease) also poses potential risks and can be unreliable in individuals with low serum ceruloplasmin concentrations (10,24). This was shown in 4 individuals in a family described previously, all of whom had low ceruloplasmin concentrations, lack of detectable incorporation of copper, and an elevation of liver copper. These characteristics initially led to the diagnosis of affected for two heterozygotes, who remained clinically asymptomatic for 20 years with no therapy (30). In initial linkage studies using esterase-D electrophoretic variants only, these two heterozygotes were still considered as patients with a recombination event (31). Copper radioisotope studies should therefore be used with caution in making a diagnosis. Thus the present diagnostic methods involve some risk, and may still be inconclusive in certain families. The development of DNA diagnosis of Wilson disease is of considerable importance. Diagnosis using DNA markers potentially provides a clear distinction between patients and heterozygotes in families with Wilson disease, as illustrated in the families presented here. In family 1, the diagnosis of 'affected' for the sister could have been established years earlier if DNA markers had been available, and before the onset of any clinical signs. If normal, extensive follow-up would not have been necessary. In a second sib the diagnosis of 'normal' was made in the presence of an equivocal sign - - infant hepatomegaly. In family 2, treatment was initiated before the diagnosis was certain and then diagnosis was difficult by the usual biochemical criteria. DNA studies confirmed the affected status. The devastating hepatic and neurological symptoms of Wilson disease can be prevented by life-long treatment with copper-chelating agents, particulary penicillamine, which are effective in achieving a negative total body copper balance. When treatment is given to presymptomatic patients, tissue damage is prevented and the overall
R.H.J. H O U W E N et al.
course is favorable (11,12). An accurate diagnosis for presymptomatic siblings of an index case is therefore essential. If a heterozygote or normal unaffected individual is treated with penicillamine, serious side-effects may result. In patients who do not have Wilson disease, e.g. rheumatoid arthritis or biliary cirrhosis, serious sideeffects occur in about 33% (32). Even for patients with Wilson disease, up to 20% experience some side-effects, half of which are serious, such as systemic lupus erythematosus or immune-complex nephritis (33). Therefore penicillamine should not be given to unaffected individuals. The use of DNA-based diagnosis requires an index case with an unequivocal diagnosis of Wilson disease, until the gene and its mutation are identified. This patient's chromosome markers serve as a reference for identification of the disease-carrying chromosomes. When the chromosomes of the sibs are studied, the chromosomes inherited from each parent can be identified, indicating whether or not the disease gene has been inherited. The sibs can be classified as patients, heterozygotes or homozygous normals. Furthermore, as our results show, with this approach 100% reliability cannot be obtained, because the diagnostic method is based on the use of flanking markers and there is always a chance of an undetected double recombinant, especially when only a few loci in the region of interest are informative in a particular individual. However, the chance of having a double recombinant can be calculated. Reliability will exceed 95% in most cases, because a large and growing number of markers on both sides of the Wilson disease gene are available. Another potential source of error in the diagnosis of Wilson disease by linked markers is the existence of more than one gene causing Wilson disease. There is some clinical evidence pointing to genetic heterogeneity, especially the observation that the affected members of a given family display similar clinical and biochemical phenotypes, while there is considerable phenotypic heterogeneity between pedigrees (24). This variation could be a result of different mutations in the same gene, from mutations at different loci, or from the action of environmental factors such as copper intake (or a combination of the above). If there were more than one locus for Wilson disease, DNA diagnosis would be impossible at present. However, current data give no support for more than one W N D locus. The gene is apparently localized on the same part of chromosome-13 in all populations studied to date: Middle-Eastern (13-15,18), Northern European (29,34), Italian (17), Chinese (35) white American (18,36) and the Canadian (Northern European ancestry) families in this report. Two studies
275
DNA DIAGNOSIS OF WILSON DISEASE
suggesting looser linkage to ESD were due to misclassification of heterozygotes as patients, as described above (31), or to broad confidence intervals associated with small sample size (29,31,34). Ultimately, the study of markers in the vicinity of the Wilson disease gene will lead to the isolation of the gene itself. This should lead to a real understanding of the pathogenesis of Wilson disease based on the gene function. It will also define the basis for heterogeneity in the clinical presentation and course of the disease. When the gene is isolated, the diagnosis of Wilson disease will also be possible by DNA methods in patients with liver or neurological diseases without a family history of this disease.
Acknowledgements We thank Prasanna Kumaranayake for technical assistance, Gail Billingsley for data verification, Drs. A. Goodridge, F.H. Anderson and W. Mackle for assistance in patient follow-up, and Drs. A.E. Bale, C.H.C.M. Buys, W. Cavenee, T. Dryja and J. Squire for DNA probes. This study was supported by the Physicians' Services Incorporated Foundation, R.H.J.H. was supported by a fellowship from the Ter Meulen Fund of the Royal Netherlands Academy of Arts and Sciences.
References I Danks DM. Disorders of copper transport. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic Basis of Inherited Disease. New York: McGraw-Hill, 1989:1411-31. 2 Hill GM, Brewer G J, Prasad AS, Hydrick CR, Hartmann DE. Treatment of Wilson's disease with zinc. I. Oral zinc therapy regimens. Hepatology 1987; 7: 522-8. 3 Hoogenraad TU, van Hattum J, van den Hamer CJA. Management of Wilson's disease with zinc sulphate. Experience in a series of 27 patients. J Neurol Sci 1987; 77: 137-46. 4 Scheinberg IH, Sternlieb I. Wilson's Disease. Philadelphia: WB Saunders, 1984. 5 Martins da Costa C, Baldwin D, Portmann B, Lolin Y, Mowat AP, Mieli-Vergani G. Value of urinary copper excretion after penicillamine challenge in the diagnosis of Wilson's disease. Hepatology 1992; 15: 609-15. 6 Evans J, Newman S, Sherlock S. Liver copper levels in intrahepatlc cholestasis of childhood. Gastroenterology 1978; 75: 875-8. 7 Smallwood RA, Wdliams HA, Rosenoer VM, Sherlock S. Lwer copper levels in liver disease: studies using neutron activation analysis. Lancet 1968; ii: 1310-3. 8 Danks DM. Copper and liver disease. Eur J Pediatr 1991; 150: 142-8. 9 Bhave SA, Pandit AN, Pradhan AM, et al. Liver disease in India. Arch Dis Child 1982: 57: 922-8. 10 Sternlieb I, Scheinberg I H. The role of radiocopper in the diagnosis of Wilson's disease. Gastroenterology 1979; 77: 138-42. I1 Sternlieb I, Scheinberg IH. Prevention of Wilson's disease in asymptomatic patients. N Engl J Med 1968; 278: 352-9. 12 Walshe JM. Diagnosis and treatment of pre-symptomatic Wilson's disease. Lancet 1988; ii: 435-7. 13 Frydman M, Bonne-Tamir B, Farrer LA, et al. Assignment of the gene for Wilson disease to chromosome 13: linkage to the esterase D locus. Proc Natl Acad Sci USA 1985; 82: 1819-21. 14 Bowcock AM, Farrer LA, Cavalh-Sforza LL et al. Mapping the Wilson disease locus to a cluster of linked polymorphic markers on chromosome 13. Am J Hum Genet 1987; 41: 27-35. 15 Bowcock AM, Farrer LA, Hebert JM, et al. Eight closely linked loci place the Wilson disease locus within 13q14-q21. Am J Hum Genet 1988; 43: 664-74. 16 Farter LA, Bonne-Tamir B, Frydman M, et al. Predicting genotypes at loci for autosomal recessive disorders using linked genetic markers: applicatton to Wilson disease. Hum Genet 1988; 79: 109-17. 17 Figus A, Lampis R, Devoto M, et al. Carrier detection and early
18
19
20
21
22 23
24
25
26
27
28
29
30
31
diagnosis of Wilson's disease by restriction fragment length polymorphism analysis. J Med Genet 1989; 26: 78-82. Farrer LA, Bowcock AM, Hebert JM et al. Predictive testing for Wilson disease using tightly linked and flanking markers. Neurology 1991; 41: 992-9. Cossu P, Pirastu M, Nucaro A, et al. Prenatal diagnosis of Wilson's disease by analysis of DNA polymorphisms. N Engl J Med 1992; 327: 57. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988; 16: 1215. Benger JC, Cox DW. Polymorphisms of the imrrmnoglobulin heavy chain delta gene and association with other constant region genes. Am J Hum Genet 1989; 45: 606-14. Church GM, Gilbert W. Genomic sequencing. Proc Natl Acad Sci USA 1984; 81: 1991-5. Williamson R, Bowcock A, Kidd K, et al. Report of the DNA committee and catalogues of cloned and mapped genes, markers formatted for PCR and DNA polymorphisms. HGM Workshop I 1. Cytogenet Cell Genet 1991; 58:1190-1832. Cox DW, Fraser FC, Sass-Kortsak A. A genetic study of Wilson's disease: evidence for heterogeneity. Am J Hum Genet 1972; 24: 646-66. Bowcock AM, Farrer LA, Hebert JM, Bale AE, Cavalli-Sforza L. A contiguous linkage map of chromosome 13q with 39 distinct loci separated on average by 5.1 centimorgans. Genomics 1991; I 1: 517-29. Houwen RHJ, Pautler SE, Barwell JA, et al. Isolation and regional localization of 25 anonymous DNA probes on a chromosome 13 hybrid panel. Cytogenet Cell Genet 1991; 57: 87-90. Ott J. Estimation of the recombination fraction in human pedigrees: efficient computation of the likelihood for human linkage studies. Am J Hum Genet 1974; 26: 588-97. Phillips M J, Poucell S, Patterson J, Valencia P. The Liver. An Atlas and Text of Ultrastructural Pathology. New York: Raven Press, 1987. Houwen RHJ, Scheffer H, te Meerman G J, van der Vlies P, Buys CHCM. Close linkage of the Wilsons disease locus to D13S12 in the chromosomal region 13q21 and not to ESD in 13q14. Hum Genet 1990; 85: 560-2. Cox DW, Billingsley GD. The application of DNA markers to the diagnosis of presymptomatic Wilson disease. In: Wetterberg L, ed. Proceedings of Genetics of Psychiatric Diseases. Wenner-Gren International Symposium, Vol. 51. Stockholm: Macmillan Press, 1988: 167-77. Cox DW, Weitkamp LR, Lewis M, Sass-Kortsak A. Linkage
276 studies in Wilson's disease (hepatolenticular degeneration). Cytogenet Cell Genet 1985; 40: 608-9. 32 James OFW. D-Penicillamine for primary biliary cirrhosis. Gut 1985; 26: 109-13. 33 Walshe JM. Wilson's disease presenting with features of hepatic dysfunction: a clinical analysis of eighty-seven patients. Q J Med New Ser 1989; 70: 253-63. 34 Scheffer H, Houwen RHJ, te Meerman GJ et al. Identification of
R.H.J. HOUWEN et al. crossovers in Wilson disease families as reference points for a genetic localization of the gene. Hum Genet 1992; 89: 607-I I. 35 Chuang L-M, Tai T-Y, Wang T-R et al. Esterase D and retinoblastoma gene loci are tightly linked to Wilson's disease in Chinese pedigrees from Taiwan. Hum Genet 1991; 87: 465-8. 36 Yuzbasiyan-Gurkan V, Brewer G J, Boerwinkle E, Venta PJ. Linkage of the Wilson disease gene to chromosome 13 in NorthAmerican pedigrees. Am J Hum Genet 1988; 42: 825-9.